Community-first. Mission complete. 🌀 Our raise on Kaito AI 🌊 Capital Launchpad did nearly 3x the target. This was only possible because of you. The dreamers who showed up and made it real. Kaito will soon share the final results and confirm allocations for those who

gEverlyn 🚀

Gm! Such a beautiful video created by our community using Everlyn. My everyday AI tools are ChatGPT, notion, gamma, and Everlyn. How about yours?

Weekend read. AI Startup Founders Tout a Winning Formula—No Booze, No Sleep, No Fun wsj.com/business/entre…

🤔 Wonder how to edit object shapes without masks or fine-tuning? Our recent work Follow-Your-Shape builds on pre-trained diffusion models to deliver a training-free, mask-free pipeline: - Trajectory Divergence Maps (TDM) to localize shape edits - Scheduled KV Injection to

Everlyn’s AI video platform is now powered by the Sui Stack ⚡ - Walrus - data layer for 5,000+ clips & onchain training - Sui - fast, reliable payments Decentralized AI, leveled up by Sui.

The wait is over at Everlyn.ai! We’ve upgraded our infra and optimized the model, greatly reducing user wait times, even as website traffic and community excitement keep growing. 🚀

Interesting! Hai Huang, Yann LeCun , and Randall Balestriero have just proposed LLM-JEPA. It's the first JEPA-style training framework for language models, bridging the gap between vision’s embedding-space objectives and NLP’s generative ones. Results: - Outperforms standard LLM

Excited to catch up with everyone to learn and share about video AI in kbw and t2049!

Building reproducible and verifiable video generation on BNB. #mvb9

Building something cool 💪

Tomorrow won’t just be another day. 🌀 It’s the beginning of something bigger. Notifications ON.

Launching 🌀

Super exciting launch, but this is only the beginning of something much larger. Our roadmap ahead is packed with stronger and faster foundation models, highly requested AI features, and deep collaborations with top universities and research labs. We’re building toward a long-term

Deploying our latest research on audio generation 🔊